Wylex 3 phase board fitting.

Presumably the hager rcbo in your screenshot is for a different job?

Just get the wylex spd kit, so much tidier.

suffolkspark said:

Presumably the hager rcbo in your screenshot is for a different job?

Just get the wylex spd kit, so much tidier.

Click to expand...

No, bollox, must have clicked the wrong link

pc1966 said:

You have to remember that the SPD is trying to divert the surge current. So if the supply goes above 500V-ish it starts to conduct rapidly and in order to dump a few kA of surge, and at for no more than a kV or so of let-through voltage, it has to have a dynamic impedance (dV/dI) of under an ohm.

But that all happens very quickly (tens of microseconds), so if the SPD's cable has any significant inductance that make the SPD look like a higher impedance, so you get less current diverted and so a bigger voltage let-through.

The SPD and its cable is a shunt across the supply, so the length of supply cable feeding it is not important, it is the length of cable between the supply L/N/E and the SPD terminals that determines how well it is able to shunt that current away (heating up internally in the process).

Click to expand...

That makes sense thanks, must have been what JW was getting at ut I couldn't remember it.

Here is an attempt to make the SPD wiring effect more obvious:

The surge on the supply is usually limited on LV networks to around 6kV open-circuit as above that something upstream will tend to flash over. While the supply cable impedance tends to 50-80 ohms for long lengths (its characteristic impedance), for 100kHz that is longer than 2km so not likely, and for shorter cables (i.e. surge closer to the SPD) it looks close to some R & L and often is as low as an ohm or so, so the surge current is often taken as around the 3kA or so region.

So your typical surge might be of the order of 3kA and 1.5kV over the SPD which is about 4.5 MW. While that is a lot of power, it is typically only for around 20us or so, so represents about 90 J of energy/heat.

Looking up zinc oxide as the typical material for a MOV surge protective device, if we take it as weight about the same as a 20p coin = 8g then it is about 0.1 mole of compound. The heat capacity is given as 40.3 J/K/mol so such a 90J surge will heat the disk up by about 23 K which is easy to survive.

Doing the same for the ~350us surge from a direct hit on to some aspect of the power system (or local ground linked to a lightning conductor), instead of the ~20us from a nearby strike inducing current, you have a disk temperature of hitting about 410 C and so melting the contacts for safe disconnecting. Next hit and your electronics is toast...

The bigger type 1(+2) SPD often use GDT (gas discharge tube) so when they fire the volts drop down to ~100V to keep the device's dissipation down by a factor of 10-15 compared to a MOV that tries to tough it out, but they have issues of mains AC current follow-through so sometimes have a lower voltage MOV in series to try and let any arc quench on the AC zero-crossing much faster, etc.

But looking at the lower circuit you can see that the L of the SPD's cables is in series with the surge current, so the volt drop along their inductance is presented to the load. Which it might not enjoy.

Thanks for the explanation, I'm not sure I understand everything but sort of get it, I've watched big clive tear down SPD devices before now and ween the gas discharge tubes inside them. I think the simplest explanation is it matters how short the cables are when they are across the mains as opposed to in-line with them.

oscar21 said:

Got loads of Wylex metal blanks, I presume they will fit in. As for surge, I'm going to fit a 3rd party one in an external enclosure, the price of the wylex one is ridiculous. Thats if it needs a SPD at all, not too sure of the commercial regulations.

Click to expand...

I have not come across many scenarios when i can omit surge protection . If Anything is going to be more expensive to repaire that the cost of the surge unit. Then I fit one. For example high level light would expensive to repaire if scissor lift was required.

oscar21 said:

Got loads of Wylex metal blanks, I presume they will fit in. As for surge, I'm going to fit a 3rd party one in an external enclosure, the price of the wylex one is ridiculous. Thats if it needs a SPD at all, not too sure of the commercial regulations.

Click to expand...

Never use metal blanks on a vertical mounted installation as I learned the hard way few months back doing a strip out on a old bupa building everything labeled up incorrectly (as I know know )removing the db cover to cheak it was isolated and one of them metal blanks slid down and landed directly across two phases sore eyes and alot of metal splash

The boards I ended up getting came with a load of plastic clip in blanks, very flimsy and wouldn't take very much to push through but obviously wylex thinks they are good enough and they must be compliant otherwise they wouldn't include them.

As for the metal blanks, the ones that you have to put in at an angle that slide about all over the place are the worst blanks I've ever tried to fit, however Wylex have come up with a new design of metal blank that clip in like the older plastic blanks did and you can fit them after the lid is on, they grip the lid really tightly as well so no chance of them falling out or sliding about.

oscar21 said:

As for the metal blanks, the ones that you have to put in at an angle that slide about all over the place are the worst blanks I've ever tried to fit, however Wylex have come up with a new design of metal blank that clip in like the older plastic blanks did and you can fit them after the lid is on, they grip the lid really tightly as well so no chance of them falling out or sliding about.

Click to expand...

They are an improvement for sure, a bit like the Axiom ones oddly enough.